The primary objective of the proposed research is to establish a computerize database for all mammographic facilities in San Francisco that can be linked to the regional population-based SEER registry for the purpose of assessing mammographic utilization, proportion of abnormal mammographic examinations, predictive value, and tests and costs associated with follow-up of abnormal mammographic results in the community (Research Plan #1). The proposed database would be based on the American College of Radiology (ACR) Breast Imaging and Reporting Data (BIRD) System or a subset of the ACR codes, depending on feasibility. Secondary objectives include assessment of the reproducibility of the ACR BIRD codes and determination of which codes best predict breast cancer (Research Plan #2) and preliminary studies of phenotypic and genotypic characteristics of screen-detected and other breast cancers (Research Plan #3). The population of the city and county of San Francisco is among the most ethnically diverse in the United States, with only 46% non-Hispanic whites, and is geographically more contained than most other urban communities. The lead institution in this proposal, UCSF, has a well- established network of clinical, epidemiologic, and basic science investigators with long-standing research interests in breast cancer; pioneering work in screening mammographic examinations has been performed here, and UCSF was recently awarded an NCI-funded SPORE grant entitled "Bay Area Breast Cancer Translational Research Program". The research platform to be developed here will be an outstanding resource for addressing issues related to mammographic screening policy, for identifying features of mammographic screening that might improve its predictive value, for biologic studies of screen-detected compared with other cancers, and for future studies of emergent screening technologies.